On 2008-03-25 18:35, fournet.arnaud wrote:
> I kindof believe there is no conclusive data
> to sort out H3e- or any Ho- in H_kw.
> Am I wrong ?
*-h3kW-o- is frequent as the second member of compounds (*X-h3kW-o-
means 'looking like X, X-like'). As the first element often ends in *-i,
the variation *-i:kWo-/*-jo:kWo- (the latter in the "breaking"
languages, the former elsewhere) demonstrates that this particular root
has *h3. And since there's no compelling semantic or formal reason to
separate it from the root of 'eye, see', the burden of the proof is on
anyone who suggests such a distinction.
Piotr